Background

A disproportionate amount of the French music performed today dates from the thirty-odd years between the start of one world war and the end of another. Maurice Ravel (1875–1937), Francis Poulenc (1899–1963), Arthur Honegger (1892–1955) and even Olivier Messiaen (1908–92) are familiar to most listeners, not forgetting that many popular works of Igor Stravinsky (1882–1971) were written in France, for France. If our sense of German music is determined above all in the nineteenth century, our idea of musical Frenchness may come rather from the early twentieth, albeit now challenged – and complemented – by our increasing awareness of the French Baroque. This accessibility (in both senses) extends back to the fin de siècle; but if it stretches to the post-Second World War years, it does so primarily through the music of composers already active before the war.

Even so, how can we sum up a period embracing works as diverse as Le boeuf sur le toit (1919) by Darius Milhaud (1892–1974) and Messiaen’s Quatuor pour la fin du temps (1941), or Stravinsky’s Symphonie de psaumes (1930) and Honegger’s Pacific 231 (1923)? Modernity and timelessness, ‘high’ and ‘low’ cultures, harmony and discord – all seem to come up against each other in this extraordinarily diverse repertoire. This was, of course, a tumultuous period, even in the context of France’s often tumultuous history. Aside from two world wars (the second of which found the country occupied and divided) and the concomitant loss of life and depletion of resources, there was social upheaval and political unrest even in times of peace. We should expect this to rub off, to some degree, on the music of the period, and particularly to be revealed in its musical culture. But should we not also imagine that a certain commonality of purpose, a few shared values, might emerge in these troubled times?

Our perspective is skewed by the passing of time. Scanning the chronology in the contemporary critic René Dumesnil’s La musique en France entre les deux guerres, 1919–1939 (1946), for example, is a mildly disconcerting experience: he lists more composers and works that are forgotten than are remembered. The same goes for Paul Landormy’s La musique française après Debussy (1943), despite its initial focus on routes to and from Les Six. This short chapter is not the place to rediscover music that, for better or worse, has fallen out of the repertoire over the intervening decades. But such books usefully remind us that ‘great works’ and ‘great composers’ are made in particular musical cultures and institutional contexts, ones whose evolution is far steadier than a history of stylistic innovation and aesthetic revolution would suppose.¹Plus ça change, plus c’est la même chose.

So I want cautiously to retain a sense of a national tradition here, while seeking to relocate some well-loved music in its context. Without underestimating the disagreement and sometimes disdain that emerged in interactions of composers and their supporters, I prefer to emphasise their often unexpected alliances and agreements. This is a historiographical decision, of course, a choice made among the narratives available. But it is one that seeks to take advantage of the limitations of space to consider this vibrant and much-discussed period in terms of continuity as well as change.

The Parade ground

The First World War, as later the Nazi occupation, is commonly imagined as a time in which musical activity must have ground to a halt. In fact, after a brief hiatus at the outbreak of conflict, cultural life resumed in a modified yet recognisable form. Questions about the seemliness of performing during the conflict faded as the war’s longevity became clear. Concerts were defended in terms of their power to raise morale, employment and taxes. By the spring of 1915, most theatres in Paris had reopened, although the most prestigious, the Opéra, did not begin performances until the end of the year. While the Colonne and Lamoureux orchestras combined their remaining personnel into one ensemble, those of the Société des Concerts formed the core of the orchestra for the patriotic Matinées Nationales held on Sundays in the huge amphitheatre of the Sorbonne.² Although Fauré was the conciliatory president of both the Société Nationale and the Société Musicale Indépendante, these composers’ societies were unable to set aside their differences: refusing to combine their efforts, they could not offer concerts at all until 1917, and then did so intermittently.³ Not for everyone, then, the truce (agreed by political parties in support of the war) known as the union sacrée.

Programming too was subject to some review. In 1916, a Ligue pour la Défense de la Musique Française, which sought to ban the performance of German music still in copyright, was established by the critic Charles Tenroc; Saint-Saëns, d’Indy and Charpentier were named among its honorary presidents. On the other hand, Fauré, Debussy, Ravel and others refused to join; recent research by Rachel Moore suggests that the league’s impact may have been minimal.⁴ The same might be said for Saint-Saëns’s infamous condemnation of the domination of German music, Germanophilie (1916), the work of a seventy-nine-year-old composer sounding some vitriolic views on the noisy battlefield of propaganda.⁵ In practice, the effect of all this on repertoire was largely limited to Wagner, whose music and ideas were sufficiently controversial to see him withdrawn from programmes.⁶ In the first season of performances, Moore has shown, German music was avoided altogether, but that restriction gradually passed as a narrative of the universality of classical masterworks re-emerged.⁷

The relationship between music and politics during the war was thus complex and contested. French orchestras went on state-sponsored tours to neutral or allied countries, performing Beethoven alongside French and Russian works; one even performed Wagner abroad.⁸ At the same time, ‘national’ French editions of German music were created to replace ‘enemy’ ones; and concerts of modern French works and Austro-German classics were often framed not just by choruses of La Marseillaise but also by patriotic speeches.⁹ Such uneasy intersections of verbal rhetoric and musical practice would be repeated many times over subsequent decades.

In this somewhat austere context, it is conventional to locate the beginning of a new, irreverent sensibility in the ballet Parade (1917) by Erik Satie (1866–1925). It is not difficult to see why. Parade famously brought together figures who were – or would go on to be – leaders in their respective fields: Pablo Picasso (1881–1973) as set and costume designer, Jean Cocteau (1889–1963) as scenarist and Léonide Massine (1896–1979) as choreographer (and dancer). It also constituted the latest succès de scandale of Serge Diaghilev’s Ballets Russes, successfully updating the company’s exotic-cum-primitivist pre-war repertoire (whose last manifestation had been Stravinsky’s Le sacre du printemps, 1913) into what would become known as the esprit nouveau.

Cocteau’s scenario itself seems to thematise questions of modern art’s relationship to its audience. The parade in question is an impromptu outdoor preview of a show, performed to drum up an audience. But the entertainers (a Chinese conjurer, an American girl and some acrobats) are so good that passers-by believe they have seen the performance gratis. Meanwhile, the managers pace anxiously, unable to persuade people to come inside the theatre. Thus Cocteau’s Parade is less a show-within-a-show (a familiar enough device) than a no-show-within-a-show: a performance of the public’s self-absorption and misunderstanding, around an empty core.¹⁰ This scenario replicated itself within the elegant Théâtre du Châtelet, where the premiere of Parade inspired the audience’s irritation and confusion.

Satie’s music did nothing to assuage concerns about the ballet. Although scored for full orchestra, it drew heavily on popular idioms of the day, seeming to blur the boundaries between ‘high’ and ‘low’ art, between what was allowed into a respectable theatre and what was left outside. To make matters worse, Cocteau had added a number of noisemakers, such as a typewriter and a foghorn; these factors combined to justify the work’s description as a ballet réaliste. Or even sur-réaliste, for this was the word coined by the celebrated modernist poet Guillaume Apollinaire (1880–1918) to describe the alliance of sets, costumes and choreography that transformed the everyday into the fantastical. Importantly, though, Apollinaire also found the music ‘astonishingly expressive … so clean-cut and so simple that it mirrors the marvelously lucid spirit of France itself’, thus tying Parade at once to the esprit nouveau and to French tradition.¹¹

Certainly, if the music of Parade is irreverent, it is a carefully constructed irreverence. The one-act ballet comprises five (later six) sections: a prelude and a coda surrounding numbers for the Chinese conjurer, the American girl and the acrobats (followed, in the final version for a 1919 revival, by reprises of each). A series of mirrors are embedded: in each section, the performers enter and exit to the same music; and the opening ‘curtain’ and managers’ entrance music is repeated in reverse order for their exit and the final curtain.¹² At the very centre, the American girl’s music derives from ragtime. This was not in itself a novelty, given that both Debussy and Satie himself had drawn on it some years earlier. But Satie went one further in Parade by parodying a specific tune, Irving Berlin’s ‘That Mysterious Rag’, whose rhythmic structure is replicated more or less exactly, while its melody is adapted, and its harmony re- or misdirected. Satie shapes the rag into a ternary form, frames it with the American girl’s entrance and exit music, and locates it in the middle of the third section of the original five.¹³ As was Satie’s practice, then, patterns and numerical relations aspire to a medieval level of intricacy and, similarly, are seen more than they are heard.

At sixes and sevens

In retrospect, this revolution, if such it was, had been heavily trailed. Satie was not of the same generation as his bright young collaborators, or of the composers of Les Six with whom he would soon be associated. Until recently, however, he had been a rather obscure figure, eking out a living as a cabaret pianist at Le Chat Noir and other venues in the bohemian quartier of Montmartre. Satie was not untrained, as is sometimes imagined: he studied lackadaisically at the Paris Conservatoire for a number of years before he was expelled. Much later, in 1905, he took himself back to school at the (still new) Schola Cantorum and worked to improve his technique in the classes of d’Indy and Albert Roussel (1869–1937), Satie’s junior by three years.

Many of Satie’s most well-known pieces date from the interim years, yet they made almost no impact until they were (re)published in the 1910s. If there is a certain naivety to such piano miniatures as Trois gnossiennes (1890–3), Pièces froides (1897) and Trois morceaux en forme de poire for piano duet (1903), with their sparse textures, odd harmonies and disappearing metres and bar lines, it is a deliberate naivety, a refusal of conventional musical codes rather than a lack of awareness of them.

Recognition as a composer was slow in coming for Satie, though his friend Debussy orchestrated two of the Gymnopédies (1888) as early as 1896. His breakthrough finally came in 1911 when, in swift succession, Ravel performed a number of Satie’s early piano pieces at a concert of the new Société Musicale Indépendante, Debussy conducted his orchestrations of the Gymnopédies at one of the Cercle Musical, and Satie began to receive favourable notices in the musical press; publications of old and new pieces soon followed. In 1913, Satie met the artist Valentine Gross (Valentine Hugo) and, through her, Lucien Vogel, who commissioned the extraordinary set of piano pieces Sports et divertissements (1914); in 1915 Gross introduced Satie to Cocteau, hence setting in motion Parade.¹⁴

Although other important works would follow (the oratorio Socrate, 1918; the ballets Mercure and Relâche, both 1924), Satie’s growing reputation over the next few years arguably owed less to his new music than it did to his social cachet and adoption as forefather by a younger generation of composers. Chief among these were the members of Les Six, Georges Auric (1899–1983), Louis Durey (1888–1979), Honegger, Milhaud, Poulenc and Germaine Tailleferre (1892–1983) – a group whose aesthetic congruity and collaborative output must not be overstated, but whose interaction and association should not be doubted.

Les Six’s nebulous origins are located among the composers who paid tribute to Satie in a series of concerts after the success of Parade: first Auric, Durey and Honegger, then Tailleferre and Poulenc, but not yet Milhaud (who did not return to Paris from Brazil until 1919). At this point, the group also included others such as Jean Roger-Ducasse (1873–1954) and Alexis Roland-Manuel (1891–1966), a loose assembly of Satie’s acolytes, including some performers and other artists, whom he referred to as ‘Les Nouveaux Jeunes’.¹⁵ Anxious to demonstrate his credentials as an impresario as well as a dramatist, Cocteau subsequently took some of the young composers under his wing and arranged for a sympathetic journalist, Henri Collet, to offer some free publicity. The group so defined collaborated on only two works, the Album des six for piano (1920) and the ballet Les mariés de la tour Eiffel (1921), by which time Durey had already deserted.¹⁶

If Cocteau had yet to receive the acclaim he desired as a writer and a dramatist, his work as a propagandist for the cause of a chic avant-garde attracted a lot of attention. In particular, his pamphlet Le coq et l’arlequin (1918) sought to define a modern aesthetic, ostensibly on the model of Satie. Here, Cocteau contrasts Satie’s linear precision with the ‘impressionist’ haze of Debussy, now cast as a Russophile Wagnerian: ‘Debussy missed his way because he fell from the German frying-pan into the Russian fire. … Satie remains intact. Hear his “Gymnopédies”, so clear in their form and melancholy feeling. Debussy orchestrates them, confuses them … The thick lightning-pierced fog of Bayreuth becomes a thin snowy mist flecked with impressionist sunshine.’¹⁷ More a series of aphorisms than a reasoned argument, Le coq is at once progressively cosmopolitan in engaging with foreign and popular music – ‘Impressionist music is outdone … by a certain American dance which I saw at the Casino de Paris’ – and oddly provincial in its insistence that ‘The music I want must be French, of France.’¹⁸ Cocteau’s dismissal of everything boche (German, i.e. Kraut) as bombastic and overblown is hardly surprising, given the date and France’s recent history of seeking to escape German influence; but it finds the self-proclaimed avant-garde writer in some curiously conservative company. He writes: ‘To defend Wagner merely because Saint-Saëns attacks him is too simple. We must cry “Down with Wagner!” together with Saint-Saëns. That requires real courage.’¹⁹

Similarly, Cocteau’s engagement with popular culture is a double-edged sword. Images of his circle ‘slumming’ to jazz at the nightclub Le Boeuf sur le Toit capture a moment in fashionable Parisian society, but Le coq et l’arlequin makes a strict division between these sources of inspiration and musicians’ own artistic outputs: ‘The music-hall, the circus, and American negro-bands, all these things fertilise an artist just as life does’, Cocteau says, but ‘These entertainments are not art. They stimulate in the same way as machinery, animals, natural scenery, or danger.’²⁰ It is a sign of weakness to derive one art from another, and Cocteau warns against it in no uncertain terms: ‘DO NOT DERIVE ART FROM ART.’²¹ Ultimately, then, popular entertainment is of interest only in as much as it helps to rid France of the perceived pretensions of German metaphysics and their realisation in overblown Romantic art: ‘what we need is a music of the earth, every-day music’.²²

As if in response to Apollinaire’s characterisation of Parade, Cocteau’s scenario of Les mariés de la tour Eiffel transformed the everyday further towards the surreal. A photographer seeks to capture wedding guests on film, but his camera instead releases the prey it caught earlier, including an ostrich, a bathing beauty and a lion (who eats a guest), while two mechanical voices issue instructions. Les mariés was premiered by the Ballets Suédois, a company set up by Rolf de Maré (1888–1964) and his star dancer and choreographer Jean Börlin (1893–1930) in ostentatious competition with the Ballets Russes. During their short existence from 1920 to 1925, the Ballets Suédois introduced, always at the neoclassical Théâtre des Champs-Élysées, any number of avant-garde works. In addition to Les mariés came ballets by individual members of Les Six: Milhaud (L’homme et son désir, 1921; La création du monde, 1923), Honegger (Skating Rink, 1922) and Tailleferre (Marchand d’oiseaux, 1923), as well, most controversially, as by Satie (Relâche, 1924).²³

An African creation myth danced to a jazzy score, Milhaud’s La création du monde is at once the most successful and the most problematic outcome of Les Six’s encounter with American popular music. While interpretations – musical, theatrical and aesthetic – are several, this piece, with its disciplined jazz fugue, is certainly not marked by Dionysian abandon. On the contrary, in both its musical form and its geometrical set and ‘dancers’, it is a work concerned, as Cocteau would have it, with measured statement and classical proportion. Nor does it leave much room for the performers’ expression. The score, even when it gestures towards improvisation, is played exactly as written, without any unconventional techniques. In Jean Börlin’s production, with scenery and costumes by the modernist artist Fernand Léger (1881–1955), the dancers were even further removed, hidden behind huge cut-outs that masked not just their faces but also their bodies. While Le sacre brings the ‘primitive’ to life (and then death), La création holds it at a cool distance.²⁴

If the circumstances of Les Six’s founding are uncertain, those of its quick demise are less clear still. Four members – Honegger, Milhaud, Poulenc and Auric – would go on to become major composers in the decades that followed, Auric primarily in music for film. Satie himself soon divided the group in two, complaining that Durey, Honegger and Tailleferre did not represent the new spirit at all and were ‘pure “impressionnistes”’.²⁵ He became associated instead with another group of young composers, known as the École d’Arcueil, after the suburb of Paris where Satie lived. All students of Charles Koechlin (1867–1950), they were Henri Cliquet (1894–1963), Roger Désormière (1898–1963), Maxime Jacob (1906–77) and Henri Sauguet (1901–89), of whom only Sauguet was ever especially celebrated as a composer. Despite all the twists and turns of the story, then, Les Six has remained a symbol of an aesthetic that was shared by few if any of its members and which they played little role in devising.

A Russian in Paris

If Wagner weighed heavily on French music across the turn of the century, the composer who caused the most soul-searching (and head-scratching) in the 1920s and 1930s was Stravinsky. Even as the conventional sketch of his career in three main periods – Russian folklore-ism, neoclassicism and serialism – has faded to reveal the common core underlying superficial difference, the stylistic shift from, say, L’oiseau de feu (1910) to the Octet (1923) is profound. The rhapsodic structure and colourful orchestration of the earlier ballet are replaced by the cold, precise tone of eight wind players who are asked not to interpret but merely to execute the notes on the page. The ‘retour à Bach’ was an efficient motto for the new aesthetic, but this music lacks both the contrapuntal complexity and the harmonic drive of the German Baroque. More germane is the stance of ‘objective’ craftsmanship and quasi-religious restraint that Stravinsky did much to cultivate, in contrast to the ‘decadent’ self-expression of Romanticism, and its extension into the self-proclaimed innovation of the avant-garde.

A great deal of scholarly energy has been expended in seeking to define twentieth-century neoclassicism – what it is and, perhaps harder, what it is not. The trouble is that composers had always modelled compositions on earlier styles or made more or less obvious reference to them in their works; the turn of the twentieth century, in particular, overflows with examples. But, as Richard Taruskin has written, ‘stylistic “retrospectivism” as such was neither a necessary component of neoclassicism or, when present, a sufficient one’.²⁶ According to his interpretation, Stravinsky’s Octet is a neoclassical piece, though it has no historical model (and, at least in the finale, it obviously draws on a recent one – ragtime).²⁷ More surprisingly, Pulcinella (1920), a ballet score that Stravinsky arranged from eighteenth-century Italian manuscripts, is not neoclassical, even though it is obviously more than a simple completion or pastiche (just ask the solo trombonist).

Pulcinella had been another commission from Diaghilev for the Ballets Russes, this time to craft a ballet from some fragments of Giovanni Battista Pergolesi (1710–36) – or, rather, from manuscripts then believed to be Pergolesi – that Diaghilev had located in Naples.²⁸ An old story tells how Stravinsky, initially reluctant, finally became so absorbed in the materials that he made an ironic reinvention of that style his own for the subsequent several decades. Taruskin argues, by contrast, that Stravinsky approached the arrangement in workmanlike fashion after some lean years; he ‘spiked’ the harmony with dissonant notes which undercut rather than conceded tonal function. Pulcinella was, in Taruskin’s words, ‘nothing to do with [Stravinsky’s] own inclinations at the time’.²⁹ This may not have been how it seemed to the notoriously unreliable composer in his later years – ‘Pulcinella was my discovery of the past, the epiphany through which the whole of my late work became possible’³⁰ – but it is a convincing argument. Ahead of Stravinsky, Diaghilev had, once again, taken the French pulse.³¹

Stravinsky’s real turning point, in the eyes of Taruskin and others, was Mavra (1922), a one-act opéra bouffe. The story, such as it is, concerns a girl who sneaks her lover into the house disguised as a maid, only for him to be discovered shaving. Although set in Russia (it is based on Pushkin), Mavra is far from the ritualistic, peasant Russia of Le sacre or Les noces (final version, 1923): the opera’s is, rather, a domesticated, bourgeois Russia, as assimilated to and by Europe. Several dance types (polonaise, polka, waltz, ragtime) imbue its seven short numbers, which are interspersed with dialogue, as do gypsy and Russian folksong, in knowing reference to Stravinsky’s own folk style. Essential to the effect, sections can simply close with a perfect cadence, in true Classical fashion, establishing formal order within the trivial drama. Where in Pulcinella Stravinsky had subtly subverted tonal function, in Mavra he plays with rather than against such logic.³²

Over the next few years, the defining works of Stravinsky’s neoclassicism (the Octet; Concerto for Piano and Winds, 1924; Sonata for Piano, 1924, etc.) tumbled out, with their short forms, precise timbres and tonal (though not unambiguous) harmonies. The aesthetic certainly overlapped with that promoted by Cocteau with and for Les Six, but it was not the same: at least in Stravinsky’s head, his was not an art of the everyday, but an art for all time, consciously striving to connect itself (and him) to the great European tradition. All the same, the fact that both Stravinsky and his music retained an unmistakable element of chic did neither his pocket nor his ego any harm.

Quite how far Stravinsky had travelled was shown when he returned to ritual of a sort in the opera-oratorio Oedipus rex (1927). A strange and arresting hybrid, Oedipus comprises a libretto by Cocteau, translated into Latin but interspersed with vernacular narration, such that the piece alternates between describing, representing and enacting the drama. The archaic language, the ritualistic repetitions of the music and the statuesque movements required of the singers combine to hold the whole in a symbolic realm – though not so securely as to rob it of dramatic power. It is as if Stravinsky were revisiting Le sacre or Les noces with a new musical language – one rooted less in Rimsky-Korsakov and Russian musical tradition and more in the Mozart, Berlioz and Verdi of their Requiems. This reconnection to the European tradition was precisely what Stravinsky sought, and thought he needed, in interwar France.³³

Older and wiser?

Satie and Stravinsky (only sixteen years apart) were far from the only composers whose careers bridged the First World War. A number of others – notably Paul Dukas (1865–1935), Koechlin, Roussel, Florent Schmitt (1870–1958) and Roger-Ducasse – would come to occupy the centre ground as composers, teachers and critics in the 1920s and 1930s. But following the deaths of Debussy in 1918, Saint-Saëns in 1921, Fauré in 1924 and Satie in 1925 – d’Indy would cling on until 1931 and Widor to 1937 – the composer best able to stand as a French challenger to the pre-eminence of Stravinsky was still Ravel.

Frequently paired with Debussy as an ‘impressionist’ (the term fits him even less well than the older composer), Ravel is in some ways better considered alongside Stravinsky, and not only because these two were closer in age. Both composers wrote a substantial part of their music for the theatre, the ballet in particular; both have an eclectic but immediately recognisable style, of extraordinary technical sophistication; and both had been members of the artistic group Les Apaches before the war, when they had even collaborated on an orchestration for the Ballets Russes. On the other hand, the two composers since that time had been moving apart. Ravel could not accept Stravinsky’s apparent volte-face in Mavra, and Stravinsky famously snubbed Ravel – impugning both his national identity and his music’s spontaneity – by calling him a ‘Swiss clockmaker’. For sure, Ravel’s compositions of the post-war years retain a more straightforward connection to his earlier works, and to those of the preceding generation, than is the case with Stravinsky; but his style, too, continued to evolve, and he was far from impervious to the charms of neoclassicism.

As Barbara Kelly has shown particularly well, a strong vein of Classicism was always found in Ravel’s music, and he naturally modelled his compositions on those of others, while also reinventing them.³⁴ Thus even such works as Ravel’s two piano concertos retain close ties to the Classical tradition, despite their obvious references to jazz, and without assuming Stravinsky’s ‘frostiness’. These popular pieces are unusual in Ravel’s (relatively small) oeuvre, which in addition to the theatre is dominated by music for salon (chamber music, songs and solo piano pieces, often later orchestrated). Both date from the turn of the 1930s, Ravel interrupting work on the Piano Concerto in G (1929–31) to write the Piano Concerto for the Left Hand (1929–30) for Paul Wittgenstein. Having taken a while to warm to it, Wittgenstein finally premiered it in Vienna in January 1932, the month in which Marguerite Long gave the first performance of the Piano Concerto in G in Paris.

One way of hearing these late pieces is as a synthesis of trends in Ravel’s music of the preceding several decades, and indeed of those in French music tout court. The single-movement Concerto for the Left Hand, for example, begins hushed with contrabassoon over cellos and basses, which is reminiscent of the two bassoons over divided basses that initiate La valse (1920) and, before that, of the lone bassoon joined by winds of Stravinsky’s Le sacre. A melancholy (bluesy?) new melody in the horn suggests Ravel’s Pavane pour une infante défunte (1899, orchestrated 1910). An orchestral tutti quickly builds, and momentarily one thinks of Daphnis et Chloé (1912) or even Debussy’s La mer (1905). As soon as the soloist enters unaccompanied, however, we are into the virtuosic, exotically harmonised piano writing of the French tradition dating back at least to Emmanuel Chabrier (1841–94). And this despite the fact that Ravel had only half as many digits available to him – a situation he addresses not by writing in a more limited register, or in fewer voices, but by rhythmically offsetting lines in such a way that an athletic hand can reach them all.³⁵

Both piano concertos reveal rather straightforwardly their debts to jazz, reminding us of Ravel’s comic opera L’enfant et les sortilèges (1925) and Violin Sonata (1927), as well as of Milhaud, Stravinsky and indeed Gershwin. To follow Ravel’s own commentary, however, this observation is less interesting in itself than in terms of the synthesis he achieves with other styles. ‘What is being written today without the influence of jazz?’ he asked: ‘It is not the only influence, however: in the concerto [in G] one also finds bass accompaniments from the time of Bach, and a melody that recalls Mozart, the Mozart of the Clarinet Quintet, which by the way is the most beautiful piece he wrote.’³⁶ Elsewhere Ravel described the work as a divertissement ‘very much in the same spirit as those of Mozart and Saint-Saëns’.³⁷ The Mozartian melody in question is the ostensibly simple one at the heart of the restrained second movement of the concerto, which is modelled on the slow movement of the Clarinet Quintet: in each case, the composers extend the melodic line to an inordinate length without once repeating themselves, or giving any inkling of the struggle involved. Ravel put it plainly: ‘That flowing phrase! … It nearly killed me!’³⁸

What is disarming about Ravel’s music of this period, then, is that it synthesises multiple sources while barely registering their incongruity and without a hint of parody. If Stravinsky’s neoclassicism came from a desire to make himself a European composer, and Les Six’s came, at least in part, from a need to remove the stain of Wagnerism or so-called Impressionism (Ravel’s included), Ravel’s was more organic, stemming from a desire to position himself in a national lineage and to model his work, albeit idiosyncratically, on the great composers of the past. It was not a historicised reinvention of earlier styles, therefore, but a progressive ‘modernization’ of them, to borrow a term from Roy Howat.³⁹ As Kelly writes: ‘Ravel drew unconsciously from his heritage, incorporating new elements into an essentially diatonic and modal framework, without overthrowing or dislocating the past.’⁴⁰

The prominence in repertoire dating from the first half of the twentieth century of ballet and various forms of ‘mixed media’ might suggest that younger composers were not as interested as Ravel in traditional forms like the concerto. Such an impression is a product less of the music actually composed during the era than of those works’ respective afterlives, however; ballet’s importance is emphasised rather than disguised by the fact that its scores are most often heard in the concert hall. For example, several composers wrote concertos that enabled them to develop second careers as soloists: this was true for works of Poulenc (notably, the Concerto for Two Pianos, 1932) and Stravinsky (Concerto for Piano and Winds, and Capriccio, 1929). These concertos are not played nearly as often as Ravel’s today, but their occasional revival enriches not only the repertoire but also our understanding of the music of the period.

Opera is another case in point. It is striking both how many operas were premiered in the interwar years and how few garnered any hold, whether nationally or internationally. The Paris Opéra fared worst of all, even under the benevolent and modernising leadership of Jacques Rouché, its director from 1914 to 1945: among the operas from this period that have stuck around, Stravinsky’s Mavra was premiered by the Ballets Russes (albeit at the Palais Garnier, home of the Paris Opéra), while Ravel’s L’enfant et les sortilèges was introduced in Monte Carlo (and then in Paris at the Opéra-Comique). The Opéra did, however, have the dubious coup of reactionary works such as Vincent d’Indy’s La légende de Saint Christophe (1915, first performed 1920). Aside from belated French premieres of foreign works (of Puccini and Strauss, for example), only Roussel’s opéra-ballet Padmâvatî (1918, first performed 1923) has limited ongoing circulation. Large-scale opera, on the historical model of the grand opéra cultivated at the Opéra in the nineteenth century, simply struggled to keep up with modern aesthetic sensibilities.⁴¹

Nevertheless, many composers were concerned, obsessed even, with adapting traditional forms to their needs and those that they perceived in French music. Milhaud, for example, who would be enduringly frustrated that he continued to be defined by a few early works, wrote six chamber symphonies (1917–23) before graduating to symphonies for full orchestra (twelve, 1939–62), as well as eighteen string quartets (1912–50, of which Nos. 14 and 15 may be combined as an octet). He even wrote a trilogy of opéras-minutes (L’enlèvement d’Europe, L’abandon d’Ariane, La délivrance de Thésée, 1927): each lasts around ten minutes and took just a day to write, and all were premiered abroad (as was his gargantuan and more or less unperformable Christophe Colomb, 1928, premiered Berlin 1930).⁴² If this might not seem a wholehearted embrace of some of the most ‘elevated’ genres of Western music, nor was it a complete rejection of them.

Honegger also wrote important works for orchestra. He, like Milhaud, however, continues to be remembered primarily for early pieces such as Pacific 231, his orchestral impression of a steam train (specifically one with two axles in front, three in the middle and one at the back). Although he protested that the title was added after the fact, it is rather hard to hear this proto-film music in any other terms; Honegger did indeed go on to compose for movies.⁴³ Yet such pictorialism is rarely felt in the symphonies (the first written in 1929–30, the other four in 1940–50), even those that carry titles, and is not typical of Honegger’s concert works. As composers matured and the bluster of the immediate post-war years faded, then, continuities with pre-war styles that had previously been hidden re-emerged.

New gods and old ones

Accounts of music in interwar France often position the 1930s as a pointed response to the 1920s, as if the Wall Street Crash of 1929 suddenly swept away frivolity and cosmopolitanism, engendering a return to tradition, religion and even reactionary politics (anticipating France’s collaboration in the Second World War). There is an element of truth in this, of course, but growing continuities with the pre-First World War era (as identified above) should not automatically indicate an about-turn on the 1920s.

On the matter of a spiritual revival, the connection between music and worship in France had never been broken, with the Schola Cantorum only the richest of several training grounds for church musicians. If Messiaen’s service for more than sixty years as organist at La Trinité is frequently sounded as a sign of his anomalous commitment to the church in a secular age, he stood in a long line of spry French organist-composers, many of whom played until their deaths (literally so in the case of Louis Vierne (1880–1937) at the console of Notre-Dame de Paris). As Nigel Simeone has recounted, César Franck served at the church of Sainte-Clotilde for more than thirty years in the late nineteenth century, his student Charles Tournemire (1870–1939) for over forty; Widor (1844–1937) spent more than sixty years at Saint-Sulpice, and his successor Marcel Dupré (1886–1971) almost forty; Maurice Duruflé (1902–86) put in forty-five years at Saint-Étienne-du-Mont, and so it goes on.⁴⁴ All these figures composed prolifically, in part as a natural outgrowth of their improvisations and church duties. Although there is not space here to survey this grand (if rather conventional) repertoire, this tradition provides an important context for composers’ engagements with religious works that is too often forgotten.

Modernist composers were increasingly drawn to sacred works, whether through renewed faith, an abstract interest in ritual or more earthly concerns. Stravinsky’s cunning dedication of his Symphonie de psaumes, a commission from Serge Koussevitzky for the fiftieth anniversary of his rich American orchestra – ‘This symphony composed to the glory of GOD is dedicated to the “Boston Symphony Orchestra”’ – rather wonderfully brings these all together. Honegger, always the most serious of Les Six, completed a number of oratorio-like works: Le roi David (1921), which made his international reputation; Judith (1925); and Jeanne d’Arc au bûcher (1935). Poulenc also wrote a lot of religious music, beginning with Litanies à la vierge noire for female chorus (1936), followed by a Mass in G major (1937) and a number of motets. This new inclination would see its fullest expression after the war in his Stabat mater (1951) and Gloria (1960), as well as his opera Dialogues des Carmélites (1956), one of the most popular of the post-war era.

The so-called return to spiritualism in 1930s France was also strongly tinged with eclecticism. If Messiaen’s Catholic faith and his dedication to the Catholic Church were not in doubt, nor were his interests in musics and practices from afar, which combine to create a distinctive sound-world. This is even more true of Messiaen’s friend André Jolivet (1905–74), whose fascination with ritual and magic was loosely informed by anthropology and channelled through his vivid imagination. Far less well known than Messiaen today, Jolivet was the only student of Edgard Varèse (1883–1965), the French-born composer who spent much of his career in the United States. Although they worked together intensely in the early 1930s, Jolivet was influenced less by Varèse’s compositional technique – the younger man wrote almost no works during this time – than by his constant experimentation with sound and search for new aural experiences.

Varèse lies in a different sense behind one of Jolivet’s first mature works, Mana for piano (1935). As the story goes, on leaving France in 1933, he gave his student a curious collection of objets d’art, which Jolivet invested with a spiritual force connecting him to his teacher.⁴⁵ Each thus spawned a movement of the suite, whose title derives from a Pacific island term (generalised in classic anthropology) for such supernatural power. Mana initiated what is sometimes described as Jolivet’s ‘magic’ or, better, ‘ritual’ period. While his freely atonal (though not serial) style obviously owes a lot to the Second Viennese School, several features of the music may be identified with French traditions of piano writing: an interest in the full timbral and textural range of the instrument; an ‘exotic’ sound-world, even given the predominantly atonal language, with pedal notes and modal passages; and a dynamic (and ritualistic) use of rhythmic stasis and propulsion. In a similar vein came Cinq incantations for flute (1936), Danse incantatoire for orchestra (1936) and Cinq danses rituelles for piano or for orchestra (1939), which collectively established Jolivet as one of the most distinctive voices of his generation.

In 1936, Jolivet and Messiaen became members of a group that seemed to strike a chord. La Jeune France comprised, in addition, Yves Baudrier (1906–88), who was its prime motivation and wrote the manifesto (but later became a composer primarily for film), and one of his teachers, Jean-Yves Daniel-Lesur (1908–2002), professor of counterpoint at the Schola Cantorum. They set out their intentions in a manifesto:

Like Les Six, La Jeune France had rather convoluted beginnings, growing in part from La Spirale, an association formed to perform and propagate new chamber music. Where La Spirale’s concerts were notably diverse (including whole concerts dedicated to contemporary music of the United States, Hungary and Germany), however, the new group’s were limited, with rare exceptions, to French composers and largely to the four members.⁴⁷

Despite friendly relations and joint concerts (which continued after the war), little actually connects the more conservative music of Baudrier and Daniel-Lesur to that of Jolivet and Messiaen, beyond a certain seriousness of purpose. In the literature, La Jeune France has sometimes been positioned in opposition to Les Six, but this is not the case: Tailleferre’s Ballade for piano and orchestra (1922) was actually heard at the inaugural concert, a ‘conscious tribute from “Les Quatre” to “Les Six”’ in the words of the Messiaen biographers Nigel Simeone and Peter Hill; and Auric, Poulenc and Honegger all wrote in support of the group.⁴⁸ In addition, the soloist in Tailleferre’s Ballade was its dedicatee Ricardo Viñes (1875–1943), an exact contemporary of Ravel’s, whose works were among the many he had premiered; and the conductor of both the first and several subsequent Jeune France concerts was Désormière, formerly of Satie’s École d’Arcueil. Here again, then, there are as many signs of collegiality and continuity among generations of composers as there are of antagonism.

Occupying time

War came to France more slowly in 1939 than it did in 1914, the official declaration in September preceding months of the so-called ‘drôle de guerre’ (phoney war). When the German offensive finally arrived in May 1940, troops swiftly outflanked the French to take Paris and led to surrender. The country was partitioned, with the north and west of France occupied by the Nazis while the south-east was left (until November 1942) to the puppet Vichy regime, located in the spa town 200 miles to the south. Parisians initially fled southwards in huge numbers, but as reports came back that life under Nazi occupation was bearable (save for certain groups), many returned, and soon a vibrant cultural life had resumed.⁴⁹

Musicians were among the many who had been called up during the phoney war and had seen active service for the few weeks of the conflict. Famously, Messiaen was one of the 1.5 million soldiers captured in June 1940 and taken to a German prisoner-of-war camp, where he spent almost a year and composed Quatuor pour la fin du temps (for violin, cello, clarinet and piano, 1940–1): not the end of time as experienced by an incarcerated soldier, Messiaen always insisted, but as signalled by the angel of the Apocalypse, to whose Revelation the quartet was an earnest response. The unusual instrumentation reflects the musicians available among fellow prisoners (the violinist Jean le Boulaire, the cellist Étienne Pasquier and the clarinettist Henri Akoka), who premiered the work with Messiaen in the camp. Across the quartet’s eight movements, the full ensemble is heard somewhat rarely (though the sixth movement is in unison throughout). This may reflect the piecemeal composition of the work as much as the peculiarity of the ensemble, however, since the movements for clarinet solo (No. 3), cello and piano (No. 5), violin and piano (No. 8) and trio sans piano (No. 4) all originated prior to the quartet’s conceptualisation as such.⁵⁰

Accounts of Quatuor pour la fin du temps have typically emphasised the remarkable conditions of its composition and premiere as the key to unlocking its meaning. In a thought-provoking discussion, however, Leslie Sprout follows Messiaen’s own first description, as well as early reviews of the work, in stressing instead its distance from the war.⁵¹ In fact, neither of the very slow duet movements, which contain the quartet’s most heart-wrenching music, originated in the camp at all: the concluding violin movement, ‘Louange à l’immortalité de Jésus’, derives from Messiaen’s Diptique for organ (1930), and the central cello movement, ‘Louange à l’éternité de Jésus’, comes, rather wonderfully, from Fêtes des belles eaux, a piece for six ondes Martenot that Messiaen wrote to accompany a water feature at the ‘Fêtes de la lumière’ of the 1937 Paris World’s Fair.⁵² Moreover, a description from the camp premiere of captured soldiers ‘divided between passionate approval and incomprehension’ sounds rather more likely than the rapt masses of Messiaen’s own later account (‘Never have I been listened to with such attention and such understanding’).⁵³

For broader French audiences at the time, Sprout argues, it was not works like Quatuor that communicated the horrors of war, but rather those like the symphonic poem Stalag IX, ou Musique d’exil (1941) by Jean Martinon (1910–76), with its folkloric interludes for flute, and particularly Jolivet’s song cycle Trois complaintes du soldat (1940), which sets his own text, written after his battalion evaded capture but lost two-thirds of its men in the process.⁵⁴ Composed for a baritone alternately representing and describing the defeated soldier, the latter piece did not entirely reject Jolivet’s modernist language, but it featured a direct form of address with which, Sprout argues, audiences could identify more easily than with Messiaen’s somewhat abstruse theological references. The second song, ‘La complainte du pont de Gien’, is also quite consonant, like a folksong partially disfigured, and reconnects with the French song tradition inherited from Fauré and Debussy.

Differing reactions to these wartime pieces were not simply responses to their musical styles. According to Sprout, ‘Critics and audiences in Paris readily accepted other modernist works as testimonials to the war, as long as they used music to confront, not escape, the harrowing current events.’⁵⁵Jolivet’s Trois complaintes were performed widely by Pierre Bernac, later orchestrated by Jolivet, and both broadcast and recorded during the war. Meanwhile, Messiaen had difficulty securing further public performances of his Quatuor, which was not finally recorded until 1957. By this point, Messiaen’s own liner notes emphasised the circumstances of the piece’s composition and its premiere in front of ‘several thousand … prisoners of all classes of society: peasants, workers, merchants, writers, doctors, priests, etc.’, as if wishing on the work a greater power to speak to ordinary people than had thus far been the case (not to mention inflating their number, since the hall in fact held fewer than 500).⁵⁶

In any case, it does no dishonour to Messiaen to observe that, within a year of his imprisonment, he was back in Paris, in a teaching position at the Paris Conservatoire, secured during a brief stay in Vichy.⁵⁷ His new post is symbolic both of the uneasy return to a form of normality as the occupation wore on, and of the exceptional circumstances: although Messiaen seems to have been in line for a position for some time, in the event he took over the harmony class of André Bloch (1873–1960), who had been removed under the Statut des juifs (Vichy’s self-imposed racial laws). More important than Messiaen’s official teaching at the Conservatoire, however, were the private classes in analysis and composition that he began to hold for a group of young composers known as ‘Les Flèches’ (The Arrows); his most celebrated graduate, Pierre Boulez (b. 1925), first attended on 8 December 1944, at which meeting Messiaen discussed Ravel’s Ma mère l’oye (1911).⁵⁸ His famous class was incorporated into the Conservatoire schedule from 1947, though officially it was in analysis and not composition.

During this period, Messiaen also completed a major exposition of his own music, his two-volume Technique de mon langage musical (Paris, 1944), with Quatuor as the prime example. He had early on devised his so-called seven modes of limited transposition (limited in the sense that, if the intervallic pattern is transposed by a semitone, one soon arrives at the same set of notes). These modes might be seen as an extension of the principle of the whole-tone and octatonic (semitone–tone alternation) scales already widely used in French and Russian music (Messiaen’s first and second modes, respectively), and like those scales remove any automatic gravitational pull (the dominant function of tonal harmony).⁵⁹ A similar interest in the limitation of possibilities and symmetrical rather than linear structures lay behind Messiaen’s principal rhythmic innovation of this period: his non-retrogradable rhythms (phrases, sometimes long, whose rhythm – though not whose pitch – is the same read backwards as forwards). Such features combine to make Messiaen’s music immediately recognisable, even after limited exposure.

Although Messiaen rarely employed twelve-tone techniques, and never did so conventionally, after the war he briefly experimented with the serialisation of rhythm, dynamics and articulation, alongside pitch, in his ‘Mode de valeurs et d’intensités’ (from Quatre études de rythme, 1950). His innovations were more important for his students, such as Boulez, who took them up and extended them, than they were for Messiaen himself, however. Indeed, Anthony Pople has noted that the spirit and in many ways even the sound of Messiaen’s music remained more closely connected to the generation of Debussy and Dukas (Messiaen’s teacher) than to Les Six, who immediately preceded him.⁶⁰ As we have seen above, the range of music of a Poulenc or a Honegger far exceeds the flippancy to which descriptions of Les Six are too often limited. Nevertheless, Messiaen’s loyalty to the music he grew up with usefully encourages us, once more, to think in terms of continuity as well as change.

So I would like to end with a work that is as canonical as any discussed here, but not in music history. La belle et la bête (1946) was only Cocteau’s second film as director (after Le sang d’un poète of 1932), though he had contributed to writing several during the war, when the French industry was surprisingly vibrant. In film studies, it is rightly celebrated as a seminal text of the fantasy genre. The music is by Georges Auric, the former member of Les Six who, back in 1918, had been the dedicatee of Cocteau’s Le coq et l’arlequin. By the point of La belle, Auric had already written more than thirty film scores, and he would finally complete well over a hundred: if this compositional mode were taken more seriously in music history, Auric’s would surely be considered a major twentieth-century voice.

The score of La belle et la bête at first sounds of a piece with countless mid-century films: it has march-like dramatic music, cymbals to the fore and soaring romantic music, with full-blooded brass and prominent harp glissandos. Gradually, though, we hear styles that are far less familiar, or rather, less familiar in this context: Auric’s ‘magic’ music has learned little from Jolivet or Messiaen, but it owes a lot to colourful French orchestral scores from Massenet to Dukas. Most striking is the sound of the beast’s spooky castle, complete with the female and male wordless choruses employed evocatively by Debussy and Ravel before the First World War. So it is hard not to hear ‘clouds, waves … and nocturnal scents’, possibly even some ‘thin snowy mist flecked with impressionist sunshine’, to turn Cocteau’s rebukes of musical impressionism in Le coq et l’harlequin against its dedicatee (and still his collaborator, some thirty years later).⁶¹

The obvious explanation for this turnaround is that all styles sooner or later become grist to the mill of the film composer, whose work relies upon familiar musical associations. In plying his new trade, Auric drew on everything he knew of his musical past. Not only Auric but also other composers who have been discussed above, notably Honegger and Baudrier, increasingly wrote for film in their later careers. Arguably, this became in the twentieth century the new compositional mainstream and an important counterpart to art music, from which it however poached constantly. A more telling way to hear La belle’s score, then, may be as part of France’s reckoning with history. If a younger generation of composers, most vocally Boulez, often defined themselves in angry renunciation of the past, an older one dealt with it more discerningly (though no less self-servingly) by choosing what to remember.

As a survey such as this one proves, a selective memory is paradoxically at once essential and antithetical to history, giving it shape at the expense of detail. Composers (and those who speak for and about them) do not simply inherit their tradition, but work to create it, crafting a past that suits their projections for the future. We should not be at all surprised if this craft and those projections change over time, least of all in the turbulent twentieth century. Yet, for all that, invented traditions are sometimes the most long-lasting and the most keenly felt. Perhaps that is one reason why this diverse repertoire, composed in times of war and peace, continues to sound to us so self-evidently – so self-confidently – French.